December 2 – March 13, 2020 (24:00 PST) | VIRTUAL HACKATHON | 18+ Only

The exploration of space continues to occur at a rapid pace. We want you to journey to Mars and use your skills to improve how rovers on the Red Planet help unlock the secrets of the universe.

Use machine learning to autonomously operate the virtual Open-Source Rover created by NASA?s Jet Propulsion Laboratory (JPL) to explore the surface of Mars.



$ 15,000.00 USD


$ 5,000.00 USD


$250 in AWS Credits to the First 100 Eligible Submissions
*credits will be rewarded after the submission deadline

The Challenge

Build and train a reinforcement learning (RL) model on AWS to autonomously drive JPL?s Open-Source Rover between given locations in a simulated Mars environment with the least amount of energy consumption and risk of damage.

You will use AWS RoboMaker for the simulation and Amazon SageMaker to build and train your RL model.

We have published all the resources you need to get started in this?GitHub repo, including a reference architecture, detailed instructions to set up your simulation and machine learning environments, a virtual model of JPL?s Open-Source Rover, and the simulated Mars world. You will submit your trained RL model along with your code.

We will evaluate all submitted models based on the weighted score and choose the top scoring submissions as winners. The highest scoring solution will also be incorporated into the JPL Open-Source Rover project for future builders to use!


Participants must use two AWS services, Amazon SageMaker?and AWS RoboMaker.

Amazon SageMaker
Amazon SageMaker helps build, train, and deploy machine learning models. SageMaker covers the entire machine learning workflow to label and prepare your data, choose an algorithm, train the model, tune and optimize it for deployment, make predictions,
and take action.


AWS RoboMaker
AWS RoboMaker helps enable your robots to stream data, navigate, communicate, comprehend, and learn.RoboMaker provides an IDE, simulation service, fleet management capabilities, and integration with other AWS services so you can build robust robotic solutions.

Provided Resources

*All the resources you need along with instructions are available in this GitHub repository.

A simulated Mars world.

A virtual model of JPL?s Open-Source Rover equipped with the necessary sensors and camera.

The cost function to be used to measure energy consumption, adjusted for the estimated risk for each of the rover?s moves in the simulated world.

The coordinates of the starting point and the destination in the simulated world.

A weighted scoring function that will be used to evaluate the models.

You are required to use the virtual model of the rover and the simulated world we provide without altering them in any fashion.

For example, you may not add a new camera or another sensor to the rover. However, you can choose any algorithm and build and train the model in any way. Once you submit your trained model, we will use it to drive the virtual rover between two random locations in a similar but random Martian terrain, and get the weighted score.


*Schedule subject to change

Competition Start

December 2nd

Competition End

March 13th @ 24:00 PST

Judging Period

March 14th – April 4th, 2020

Winners Announced

Date and Venue to be announced

Judging Criteria

We will publish a formula that will be used to evaluate your reinforcement learning (RL) model. The formula provides a single score for your model and will be based on the rover?s ability to:

  • Reach destination points?
  • Reach destination points in the least amount of time?
  • Reach destination points with least amount of power consumed?
  • Reach destination points with the correct rover orientation
  • Reach destination points with the lowest total acceleration / ?jerk? to the body of the rover

To be eligible for judging, you must follow the reference architecture?provided in this?GitHub repository.

We utilize a hackathon event facilitation platform.

Submission Requirements

To enter the competition, you will need to submit your path_optimizer ros package along with the RL model you are using in the simulation. You can package this up as a tar or zip file and upload it via the portal.

Your application:

  • Must use the simulation packages (rover_description and rover_gazebo) provided with the sample application kit.
  • Must not use any pre-created maps for navigation. The robot must run? autonomously in the Gazebo world provided.
  • Must use the RL environment provided.


Who can participate?

All developers, designers, and entrepreneurs over the age of 18.

How are teams formed?

Teams can be created in advance using the AWS / JPL Open-Source Rover Challenge Slack Workspace, in the #teambuilding channel. Teams may be comprised of one to five individuals.?

How will my project be judged?

We utilize as a hackathon event facilitation platform.

Who Owns the IP?



Participants own the rights to their projects they create at the Hackathon, but entrants agree to submit their entry to the contest under the Apache-2.0 open source license, as the winning solution will be incorporated into the JPL Open-Source Rover project. As such, all entrants confirm that they have the right to submit their technology subject to the terms of the Apache-2.0 license. Your team may create a prototype using data and/or API?s provided by The Sponsor and/or its Partner. This could mean that you or a teammate created the Technology, acquired ownership of the Technology from a third party, or may rely on binding written statements by the third party that owns or has the right to license the Technology indicating that you (or members of the public generally) are authorized to use that Technology in the manner you intend to use it. Your team will be disqualified if The Sponsor has any reason to believe that your team has violated the terms of this paragraph. You should consult with appropriate advisors or legal counsel if you have any doubt as to whether you are meeting the requirements of this paragraph. ?Technology? means, without limitation, content (including pictorial, audio and audio-visual content), code, specifications, technical information, algorithms, images, design, art, music, graphics, SFX, data, and any other information or materials protected by any intellectual property right. Your team may bring to the Hackathon any pre- developed or licensed Technology that you plan to use in connection with your prototype, provided that such Technology meets the requirements of this paragraph. By participating in the Hackathon, you will receive access to certain proprietary software, APIs, and/or other copyrighted materials, including pictorial, audio, video and/or audio-visual content owned by the Sponsor or its affiliates, partners or licensors.